Research map / Graph learning / GNN architectures
Gauge-Equivariant Graph Networks via Self-Interference Cancellation
Yoonhyuk Choi, Jiho Choi, Jiwoo Kang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add self-interference cancellation (SIC) (ablation) | GESC · Cora | accuracy (%): 82.4 -> 84.9 | Improved |
| add self-interference cancellation (SIC) (ablation) | GESC · Chameleon | accuracy (%): 60.7 -> 65.0 | Improved |
| add self-interference cancellation (SIC) (ablation) | GESC · Texas | accuracy (%): 68.2 -> 74.5 | Improved |
| add residual gating (RG) (ablation) | GESC · Citeseer | accuracy (%): 70.0 -> 72.1 | Improved |
| add residual gating (RG) (ablation) | GESC · Chameleon | accuracy (%): 62.6 -> 65.0 | Improved |
| add gauge-equivariant transport (GET) (ablation) | GESC · Citeseer | accuracy (%): 69.2 -> 72.1 | Improved |
| add gauge-equivariant transport (GET) (ablation) | GESC · Chameleon | accuracy (%): 61.8 -> 65.0 | Improved |
| add gauge-equivariant transport (GET) (ablation) | GESC · Texas | accuracy (%): 70.7 -> 74.5 | Improved |
| use complex-valued rather than real-valued representations (ablation) | GESC · Citeseer | accuracy (%): 68.6 -> 72.1 | Improved |
| use complex-valued rather than real-valued representations (ablation) | GESC · Chameleon | accuracy (%): 60.9 -> 65.0 | Improved |
| add residual gating (RG) (ablation) | GESC · Cora | accuracy (%): 83.8 -> 84.9 | Improved |
| stack up to 12 layers | GESC vs GCN/GAT · Cora and Chameleon | accuracy (%): GCN and GAT suffer progressive degradation with depth; GESC maintains consistently higher accuracy | Improved |
Closest papers by meaning
- GIST: Gauge-Invariant Spectral Transformers for Scalable Graph Neural Operators (NeurIPS 2026 · similarity 0.85)
- Beyond Oversquashing: Understanding Signal Propagation in GNNs Via Observables (NeurIPS 2026 · similarity 0.85)
- Higher-Order Graphon Neural Networks: Approximation and Cut Distance (ICLR 2025 · similarity 0.85)
- Are High-Degree Representations Really Unnecessary in Equivariant Graph Neural Networks? (NeurIPS 2024 · similarity 0.85)
- Accurate and Scalable Graph Neural Networks via Message Invariance (ICLR 2025 · similarity 0.85)
- On the Expressive Power of Spectral Invariant Graph Neural Networks (ICML 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
